A positive association between socio-economic status (SES) and good mental health outcomes has long been observed in many contexts. Active research continues to try to tease out answers about the direction of causality in this relationship, and about the causal mechanisms at play. This research proposes to extend this literature by exploiting the longitudinal structure of data in the 1979 cohort of the National Longitudinal Survey of Youth (NLSY79). The research will provide insight into causality by testing hypotheses about temporal sequencing in the relationship between depression outcomes and income, controlling for other socio-demographic covariates, such as age, race, education and gender. This analysis will also permit the researcher to test whether the effects of economic status upon depression outcomes and vice-versa are stronger for minority sub-populations. The independent effect of local income inequality will also be tested. The proposed training component of this grant period is designed to build on the investigator's strengths as an economist by providing further exposure to recent research in mental health practice, etiology and epidemiology. This training will take the forms of courses, conferring with the primary mentors and consultants, and participation in relevant workshops in the Psychiatry Department and elsewhere in the University. Finally, the last 18 months of the grant will be devoted to developing a new research protocol on the role of time preference (i.e., impatience or delay of gratification) in determining patient demand for mental health services.